Research Center, Florence, SC 29501 USA c Department of Chemistry, Clemson University, Clemson,. SC 29634 USA. Despite considerable efforts in ...
notes
Comparative Investigation of Fourier Transform Infrared (FT-IR) Spectroscopy and X-ray Diffraction (XRD) in the Determination of Cotton Fiber Crystallinity Yongliang Liu,a,* Devron Thibodeaux,a, Gary Gamble,a,à Philip Bauer,b Don VanDerveerc a
USDA, ARS, Cotton Quality Research Station, P.O. Box 792, Clemson, SC 29633 USA b USDA, ARS, Coastal Plain Soil, Water, and Plant Research Center, Florence, SC 29501 USA c Department of Chemistry, Clemson University, Clemson, SC 29634 USA
Despite considerable efforts in developing curve-fitting protocols to evaluate the crystallinity index (CI) from X-ray diffraction (XRD) measurements, in its present state XRD can only provide a qualitative or semi-quantitative assessment of the amounts of crystalline or amorphous fraction in a sample. The greatest barrier to establishing quantitative XRD is the lack of appropriate cellulose standards, which are needed to calibrate the XRD measurements. In practice, samples with known CI are very difficult to prepare or determine. In a previous study,14 we reported the development of a simple algorithm for determining fiber crystallinity information from Fourier transform infrared (FT-IR) spectroscopy. Hence, in this study we not only compared the fiber crystallinity information between FT-IR and XRD measurements, by developing a simple XRD algorithm in place of a time-consuming and subjective curve-fitting process, but we also suggested a direct way of determining cotton cellulose CI by calibrating XRD with the use of CIIR as references. Index Headings: Fourier transform infrared spectroscopy; FT-IR spectroscopy; X-ray diffraction; XRD; Cellulose; Cotton fiber; Crystallinity index.
INTRODUCTION Cotton fiber consists of natural cellulose I (b 1!4 linked glucose residues) and its quantity increases greatly with the Received 31 January 2012; accepted 2 May 2012. * Author to whom correspondence should be sent. E-mail: yongliang.liu@ ars.usda.gov. Current address: USDA, ARS, Cotton Structure & Quality Research Unit, 1100 Robert E. Lee Blvd. New Orleans, LA 70124. Current address: 103 Long View Ct., Pickens, SC. à Current address: USDA, ARS, Quality and Safety Assessment Research Unit, Athens, GA. Mention of a product or specific equipment does not constitute a guarantee or warranty by the U.S. Department of Agriculture and does not imply its approval to the exclusion of other products that may also be suitable. DOI: 10.1366/12-06611
Volume 66, Number 8, 2012
stages of fiber growth.1–3 Hydroxyl groups in cellulose are involved in a number of intra- and inter-molecular hydrogen bonds, which result in ordered crystalline and disordered (amorphous) regions. A parameter termed the crystallinity index (CI) has been used to describe the relative amount of crystalline portion in cellulose, based on the traditional twophase cellulose model (crystalline vs. amorphous).4–6 The crystalline structure of cotton fibers has been studied predominantly using the X-ray diffraction (XRD) technique.4–7 To calculate the CI of cotton cellulose, two different methods have been used. The first one was developed by Segal and coworkers,7 who estimated the CIs from the simple formula 100(1 Iam/I002) on the basis of two XRD peaks at 2h = 22.78 and 188. As this method was proposed for empirical measurements to allow rapid comparison of cellulose samples, it should not be used as a method for calculating the amount of crystalline fraction in a sample.8 One of the biggest concerns is that the Iam value is significantly underestimated, resulting in an overestimation of the CI. The second approach, also the most practiced, is a deconvolution method, in which individual crystalline peaks were extracted by a curve-fitting process from the diffraction intensity profiles.4–6,8 It requires computing software to separate amorphous and crystalline contributions to the XRD intensities. During the curve-fitting protocol, a few assumptions have to be made, such as the shape and number of peaks as well as the deconvolution functions. Then, the CI was calculated from the ratio of the area of all crystalline peaks to the total area. An important hypothesis for this analysis is that increased amorphous contribution is the main contributor to peak broadening. In addition to amorphous content, there are other intrinsic factors that influence peak broadening, such as crystallite size. It might be possible to deconvolve these contributions with well-behaved samples that can be resolved into many narrow diffraction peaks over a significant range of 2h. Unfortunately, cellulose peaks are very broad, overlapped, and not well resolved. It is generally accepted in the cellulose community that peak broadening is due to the amorphous cellulose. Meanwhile, crystallite size is an equally important issue for peak broadening and information about average crystallite size has been assessed from XRD.4,5 In order to reduce the effect of the amorphous fraction on CI readings, subtracting the amorphous contribution to XRD intensities using an amorphous standard was outlined by Ruland.9 However, the challenge is to select an amorphous standard that is similar to the amorphous component in the sample. Traditionally, ball-milled cellulose has been utilized for such a purpose.7 Other analytical means, including solid-state 13C nuclear magnetic resonance (NMR)8,10 and infrared11,12 and Raman13 spectroscopy, have been explored to investigate the structure and CI of various celluloses. However, there are few reports available that relate the CIs from XRD technique to those from NMR, Raman, or IR tools. In a recent study, Park et al.8 compared the CIs of eight different celluloses between XRD and 13C NMR and found that the simplest XRD ratio method
0003-7028/12/6608-0983$2.00/0 Ó 2012 Society for Applied Spectroscopy
APPLIED SPECTROSCOPY
983
produced significantly higher crystallinity values than did the other methods. In general, XRD determination of cellulose CI provides a qualitative or semi-quantitative evaluation of the amounts of either amorphous or crystalline components in a sample.8 Development of a reliable quantitative cellulose CI protocol is desired by utilizing appropriate cellulose standards to calibrate or validate the XRD measurement. Apparently, these absolute standards are not easy to prepare or determine. In the latest Fourier transform infrared (FT-IR) study on immature and mature cotton fibers,14 we proposed a ratio algorithm R2 by linking it with the relative amount of Ib to Ia crystal forms, represented by the respective IR bands at 708 and 730 cm1. Applying this algorithm to the data set of 201 immature and 201 mature seed bolls suggested an R2 range of 1.54 to 2.37 for immature fibers and 2.10 to 3.17 for mature samples. By setting an R2 threshold value at 2.24, only six immature fibers and ten mature samples were misclassified, yielding an overall identification accuracy of 96%. Despite the relative weakness of the intensities at 730 and 708 cm1, acceptable classification efficiency indicated the separation power of the R2 algorithm, which is anticipated because these values should be associated with cotton growth or development. A need for rapid, accurate, and routine technologies to assess cotton crystallinity has been a subject of continuous interest for many years. Undoubtedly, the well-defined XRD method is mostly acceptable. However, its measurements are destructive (i.e., fibers must be cut or milled) and time consuming (;60 min/sample), and notably, the degree of CI depends on experimental parameters and subjective curve-fitting steps. This is one of many factors making it difficult to compare the CIs from different research groups. For example, Sarna and Wlochowicz15 reported a CI range of 73% to 80% for Turkmenistan and Uzbekistan native cottons, which was higher than the range of 53% to 69% for Egyptian cottons,16 57–64% for Saudi Arabian cottons,17 and 30–66% for upland cottons at growth stages of 21 to 60 days post-anthesis (DPAs).4,5 The objectives of this study were (1) to determine the cotton fiber CIs at single boll level from FT-IR measurement (CIIR), (2) to propose a simple ratio algorithm for XRD crystalline information, (3) to correlate the XRD reading with known CIIR, and (4) to convert the XRD reading into respective CI values (CIXRD). In addition, due to the micro-sampling feature of FTIR spectral acquisition, it is possible to look into the differences of CIIR between various portions within a boll, which is impossible through current crystalline measurements.
MATERIALS AND METHODS Seed Cotton Bolls. Seed and dried locules (one of three to five segmented compartments in a boll) were collected from ten cotton bolls at varying developmental stages of 21 through 60 DPAs. Two commercially available cultivars representing two varieties of Deltapine (DP) 0949B2RF and Phytogen (PHY) 565WR were grown in Florence, South Carolina, in 2010 crop year. Unlike the water rinsing of seed fibers described by Hu and Hsieh,5 these bolls received no pretreatment. However, they were well conditioned at a constant relative humidity of 65% and temperature of 22 6 2 8C for at least 48 hours prior to FT-IR and XRD measurement, FT-IR Spectral Collection. All spectra were collected with an FTS 3000MX FT-IR spectrometer (Varian Instruments, Randolph, MA) equipped with a ceramic source, KBr beam
984
Volume 66, Number 8, 2012
splitter, and deuterated triglycine sulfate (DTGS) detector. The attenuated total reflection (ATR) sampling device utilized a DuraSamplIR single-pass diamond-coated internal reflection accessory (Smiths Detection, Danbury, CT), and a consistent contact pressure was applied by way of a stainless steel rod and an electronic load display. Ten measurements at one of two locations (either medium or tip positions) of one locule were collected over the range of 4000–600 cm1 at 4 cm1 and 32 co-added scans. Caution was taken to avoid both seed and visible trash portions in samples. All spectra were given in absorbance units and no ATR correction was applied. Following the import to Grams/AI software (V.7, Thermo Galactic, Salem, NH), the spectra were smoothed with a Savitzky–Golay function (polynomial = 2 and points = 11). The data set was then loaded into Microsoft Excel 2000 to execute simple algorithm analysis. XRD Acquisition. XRD data were obtained using a Rigaku Ultima IV Diffractometer (Rigaku, The Woodlands, TX). It was equipped with a copper anode sealed tube using an accelerating voltage of 40 kV with a tube current of 40 mA. The geniometer scanned a 2h range between 58 and 408 with a scan rate of 0.78/min. Ten cotton fiber samples from DP and PHY varieties were prepared by Wiley milling into a power that was passed through a 20-mesh screen. The cut fibers were placed firmly on an aluminum holder (25 mm diameter 3 2 mm deep) and then mounted at the center of the goniometer circle for scanning. All data were imported to Grams/AI software and smoothed with a Savitzky–Golay function (polynomial = 2 and points = 15) before exporting into Microsoft Excel 2000 to execute simple algorithm analysis.
RESULTS AND DISCUSSION FT-IR and CIIR. Figure 1 shows the representative FT-IR spectra of developing cotton fibers at 20, 34, and 60 DPAs of DP variety in the 3600–600 cm1 region. Band assignments for the cotton fibers have been studied in some detail.1,2,18 According to the band assignments, features between 3600 and 2750 cm1 are from the O–H and C–H stretching vibrations. A broad band centered at 1640 cm1 is attributed to the OH bending mode of water. Bands in the region of 1500–1200 cm1 are mixtures of CH2 deformations and C–O–H bending vibrations, and a number of bands in the 1200–900 cm1 region are assignable to coupling modes of C–O and C–C vibrations; also, the bands between 800 and 700 cm1 are likely attributable to two crystal forms (Ia and Ib) of cotton cellulose. Comparison of the spectra reveals some common bands, indicating that the dominant FT-IR bands arise from the major common chemical component in cottons, cellulose. Given the significant structural and compositional variations among developing cottons, distinctive spectral differences occur in several wavenumbers, for example, pronounced shoulders at 2916, 1003, and 980 cm1. In the previous report regarding the spectral distinctions between immature and mature fibers,14 we developed the threeband ratio algorithm (R2) given in Eq. 1 to assess the crystalline information: R2 ¼ ðI708 I800 Þ=ðI730 I800 Þ
ð1Þ
Next, we could convert the R2 values in Eq. 1 to CIIR by Eq. 2: CIIR ð%Þ ¼ ðR2 R2;sm Þ=ðR2;lr R2;sm Þ100
ð2Þ
FIG. 2. Typical XRD pattern of cotton fibers at 20 (dotted line), 34 (solid line), and 60 (dashed line) DPAs.
FIG. 1. Representative FT-IR spectra of cotton fibers at 20 (dotted line), 34 (solid line), and 60 (dashed line) DPAs.
in which CIIR represents the CI through the FT-IR measurement; and R2, R2,lr, and R2,sm are the R2 values for unknown sample, the largest R2, and the smallest R2 values, respectively. If the most immature and mature fibers are assigned the CIIR value of 0.0% and 100.0%, respectively, the corresponding R2 values should be the smallest and largest ones in Fig. 3.14 Hence, the R2,sm and R2,lr were determined to be 1.40 and 3.40, respectively, which were approximately 0.20 units smaller and larger than those observed in Fig. 3.14 Both Eqs. 1 and 2 were applied to the FT-IR spectral set, and the resulting CIIR values from the average of ten medium and ten tip portions of the respective locules are outlined in parenthesis in Table I. Given the limited number of cotton bolls, it is inappropriate to draw any solid conclusions about the two varieties, because fiber growth depends on environmental conditions, such as soil and climate. As expected, Table I reveals that CIIR increases from shorter to longer DPAs for each variety. In general, the CIIR readings are in good agreement with Hu and Hsieh’s XRD approaches for developing cottons.4,5 For example, DP variety matches well with Maxxa cottons5 while the PHY fibers are close to Acala ones.4 XRD and CIXRD. The normalized and smoothed XRD pattern of cotton fibers at developmental stages of 20, 34, and 60 DPAs are shown in Fig. 2. For the comparison, the XRD readings were normalized by dividing the intensities with a total integrated area in the 2h region of 10–308. The intact samples were scanned for FT-IR spectra first, then they were cut into powders that were scanned for XRD characterization. The peaks located near 2h = 14.9, 16.7, 22.8, and 34.68, which are the characteristics of 101, 101, 002, and 040 reflections of cellulose I,4,5 were measured. With fiber growth, major peak intensity near 22.88 increases and its peak half-height width decreases, both of which suggest that the crystallinity and the
crystallite dimension increase with secondary wall cellulose biosynthesis. Similarly, we proposed the 4-band ratio (R3) in Eq. 3 to relate the XRD measurement to the crystalline information: R3 ¼ ðI22:8 þ I20:5 Þ=ðI14:9 þ I16:5 Þ
ð3Þ
where R3 is indicative of XRD crystalline information and I14.9, I16.5, I20.5, and I22.8 are each a three-point average of the intensity values at individual 2h position. In anticipation, the R3 values increase with DPAs (not shown). Relationship of CIIR versus CIXRD. In order to compare the crystallinity readings from FT-IR and XRD measurements,
FIG. 3. Plot of CIIR from FT-IR procedure vs. CIXRD from XRD measurements.
APPLIED SPECTROSCOPY
985
TABLE I. Comparison of cotton fiber CI values (%) for diverse varieties from different groups. Varieties
DPa
PHYa
Maxxa5
DPA = DPA = DPA = DPA = DPA = DPA = DPA = DPA = DPA = DPA = DPA = Others
31.2 (27.7)
31.9 (33.6)
31
a
20 21 25 27 30 34 36 42 48 53 60
Acala4
Turk./Uzb.15
Egyp.16
Saudi17
73–80
53–69
57–64
30 54 45.6 (44.6)
39.7 (41.6)
45 64
63.0 (63.0)
48.2 (53.4)
55 66
68.0 (66.0) 56 72.9 (74.6) 72.4 (70.0)
63.2 (61.2) 58
From this study; numbers in parenthesis were from FT-IR measurements (CIIR).
Fig. 3 shows the plot of CIIR from FT-IR spectra against R3 from the XRD scan. The regression line reveals a higher correlation of R2 . 0.90, which is within the expectation since the two methods should be consistent in reflecting the compositional and structural changes among developmental cottons. Furthermore, the R3 values could be converted into respective CIXRD readings in the range of 0.0% to 100.0% by Eq. 4, which is given as a secondary vertical axis in Fig. 3: CIXRD ð%Þ ¼ 100ðR3 1:317Þ=0:91
ð4Þ
Table I also compiles the CIXRD readings in this study. They range from 31.2% to 72.9% for DP cottons and 31.9% to 63.2% for the PHY variety. In general, the CIXRD readings are in good agreement with Hu and Hsieh’s values for developing cottons.4,5 When making a comparison with cottons from commercial bales, these observations are between the reported CI range of 73% to 80% (on Turkmenistan and Uzbekistan native cottons) and those of either 53–69% (on Egyptian cottons) or 57–64% (on Saudi Arabian cottons). Hence, it might suggest that the strategy reported here to assess cotton crystallinity is appropriate and reasonable.
CONCLUSION Major XRD peaks were used to develop the simple ratio algorithm and then the ratios were compared with the CIIR from FT-IR analysis. Within limited samples, strong correlation between CIIR and R3 (or CIXRD) suggested the equivalence and effectiveness of two separated measurements in crystallinity characterization. In other words, samples with easy and rapidly determinable CIIR values could be used to calibrate the XRD measurement. Also, the CI readings from this work were consistent with those from traditional, time-consuming, and subjective protocols used by other diverse researchers. On the basis of practical implementation of the FT-IR procedure for the determination of fiber crystallinity, this study described a simple procedure to acquire the ATR-device-based FT-IR spectra and to analyze the spectra. This protocol avoids the need to perform any pretreatment of the cotton fibers and is also advantageous in using small amount of fiber (as little as 0.5 mg) and in requiring only a short time (less than 2 min) for sample loading, spectral acquisition, and subsequent result reporting.
986
Volume 66, Number 8, 2012
1. Y.-L. Hsieh. ‘‘Chemical Structure and Properties of Cotton’’. In: S. Gordon, Y.-L. Hsieh, editors. Cotton: Science and Technology. Cambridge, UK: Woodhead Publishing Limited, 2007. Pp. 3-34. 2. N. Abidi, L. Cabrales, E. Hequet. ‘‘Fourier Transform Infrared Spectroscopic Approach to the Study of the Secondary Cell Wall Development in Cotton Fiber’’. Cellulose. 2010. 17: 309-320. 3. N. Abidi, E. Hequet, L. Cabrales. ‘‘Changes in Sugar Composition and Cellulose Content during the Secondary Cell Wall Biogenesis in Cotton Fibers’’. Cellulose. 2010. 17: 153-160. 4. Y.-L. Hsieh, X.-P. Hu, A. Nguyen. ‘‘Strength and Crystalline Structure of Developing Acala Cotton’’. Textile Res. J. 1997. 67(7): 529-536. 5. X.-P. Hu, Y.-L. Hsieh. ‘‘Effects of Dehydration on the Crystalline Structure and Strength of Developing Cotton Fibers’’. Textile Res. J. 2001. 71(3): 231-239. 6. R. Jenkins, R. Snyder. ‘‘Introduction to X-ray Powder Diffractometry’’. In: Chemical Analysis. New York, USA: John Wiley and Sons, Inc., 1996. Volume 138. 7. L. Segal, J.J. Creely, A.E. Martin, C.M. Conrad. ‘‘An Empirical Method for Estimating the Degree of Crystallinity of Native Cellulose using the Xray Diffractometer’’. Textile Res. J. 1962. 29: 786-794. 8. S. Park, J.O. Baker, M.E. Himmel, P.A. Parilla, D.K. Johnson. ‘‘Cellulose Crystallinity Index: Measurement Techniques and Their Impact on Interpreting Cellulase Performance’’. Biotechnol. Biofuels. 2010. 3: 1-10. 9. W. Ruland. ‘‘X-ray Determination of Crystallinity and Diffuse Disorder Scattering’’. Acta Cryst. 1961. 14: 1180-1185. 10. P.T. Larsson, K. Wickholm, T. Iversen. ‘‘A CP/MAS C-13 NMR Investigation of Molecular Ordering in Celluloses’’. Carbohydr. Res. 1997. 302: 19-25. 11. Y. Kataoka, T. Kondo. ‘‘FT-IR Microscopic Analysis of Changing Cellulose Crystalline Structure during Wood Cell Wall Formation’’. Macromolecules. 1998. 31: 760-764. 12. M. Akerholm, B. Hinterstoisser, L. Salmen. ‘‘Characterization of the Crystalline Structure of Cellulose using Static and Dynamic FT-IR Spectroscopy’’. Carbohydr. Res. 2004. 339: 569-578. 13. K. Schenzel, S. Fischer, E. Brendler. ‘‘New Method for Determining the Degree of Cellulose I Crystallinity by Means of FT Raman Spectroscopy’’. Cellulose. 2005. 12: 223-231. 14. Y. Liu, D. Thibodeaux, G. Gamble. ‘‘Development of FTIR Spectroscopy in Direct, Non-Destructive, and Rapid Determination of Cotton Fiber Maturity’’. Textile Res. J. 2011. 81(15): 1559-1567. 15. E. Sarna, A. Wlochowicz. ‘‘Changes in the Superstructure and Surface Morphology of Cotton during Spinning’’. Textile Res. J. 2003. 73(3): 216220. 16. A.M. Hindeleh. ‘‘Crystallinity, Crystallite Size, and Physical Properties of Native Egyptian Arabian Cotton’’. Textile Res. J. 1980. 50(11): 667-674. 17. B.A. Khalifa, N. Abdel-Zaher, F.S. Shoukr. ‘‘Crystalline Character of Native and Chemically Treated Saudi Arabian Cotton Fibers’’. Textile Res. J. 1991. 61(10): 602-608. 18. Y. Liu, D. Thibodeaux, G. Gamble. ‘‘Characterization of Attenuated Total Reflection Infrared Spectral Intensity Variations of Immature and Mature Cotton Fibers by Two-Dimensional Correlation Analysis’’. Appl. Spectrosc. 2012. 66(2): 198-207.
